Google Opens SynthID Detector Worldwide to Check AI-Generated Media

Google has made its SynthID Detector available worldwide in English, widening access to a tool that was previously introduced for media professionals. Announced on October 7, 2026, the change gives everyday users another way to examine whether an image, video or audio file carries a watermark associated with AI-generated media.

That sounds simple, but there is an important distinction: SynthID is a watermark detector, not a universal truth machine. A positive finding can provide useful provenance information for supported content. A negative finding does not prove that something was filmed, photographed or recorded by a person.

What changed on October 7

Google says the expanded detector can check media created using participating AI systems from Google and partners. The company names OpenAI, NVIDIA and Kakao among its partners, with Apple support described as coming soon. The service is now open globally in English rather than being limited to a smaller professional group.

The tool works with supported image, video and audio files. Google's underlying SynthID technology embeds signals into AI-generated media that are designed to remain difficult for people to notice. The detector looks for those signals; it does not infer authenticity from how realistic a face, landscape or voice happens to look.

What a result really means

Imagine receiving a spectacular storm photograph in a group chat. A watermark match could help you establish that the image was generated or modified with a participating AI tool. But if the detector reports no match, there are several possibilities: the image may be genuine, it may have come from a nonparticipating generator, or its watermark may be absent or undetectable.

That is why a good verification habit still includes checking the original source, publication date, context and independent reporting. The detector adds evidence; it does not replace judgment. This matters in the Philippines, where viral screenshots, edited clips and forwarded voice notes can spread quickly through messaging apps.

Why Google is expanding it

Google says SynthID has already been used to watermark more than 180 billion images and videos, as well as 240,000 years of audio content. Those figures describe the scale of watermarking across participating products, not the number of fake posts caught online.

Google also offers verification features in Search, the Gemini app and Chrome. Its latest announcement positions the standalone detector as an additional option for people who need to check a particular file rather than rely on a social platform's labeling.

How to try it

Visit the official SynthID Detector and follow the site's instructions for supported files. Avoid uploading confidential documents, personal recordings or sensitive images unless you are comfortable with the service's terms and privacy practices. Results should be interpreted as one signal in a broader verification process.

The takeaway: this is a useful new public tool for checking some AI-generated media, but it is not a certificate of reality. That limitation is as important as the launch itself.

Source: Google's October 7 announcement.